search

labsyspharm / ashlar

ASHLAR: Alignment by Simultaneous Harmonization of Layer/Adjacency Registration
https://labsyspharm.github.io/ashlar/
MIT License
126 stars 42 forks source link

using ASHLAR for registering whole slide imaging data #105

Open alcrevenna opened 2 years ago

alcrevenna commented 2 years ago

Dear @Yu-AnChen,

based on your post below I just want to add another interested person in registering data from slide scanners. In particular we have an olympus vs200. Output format is vsi but can be converted to tif. Happy to know more if progress is made on this, or to provide data to test it. Alvaro

Thanks, having the example images would be helpful! ASHLAR is designed for stitching and registering imaging data that are not stitched. However, we recently took the idea in ASHLAR and started another module that hopefully works with pre-stitched images. Do you know what scanner was used and what type of files can it export? Like .tif .czi?

Originally posted by @Yu-AnChen in https://github.com/labsyspharm/ashlar/issues/99#issuecomment-983116495

Yu-AnChen commented 2 years ago

Hi @alcrevenna,

I looked around a bit trying to find a good way to read .vsi/.ets files and saw that hopefully the support from openslide will come. AICSimageio can use BioFormats to read .vsi/.ets but it doesn't seem to get the right image size for other pyramid levels except for the full resolution level. For now, the options would be

  1. Exporting/converting vsi to a disk storage format that can be read as dask array, such as tiff, zarr, to save RAM usage during the registration steps
  2. Or read the vsi data into RAM (using AICSimageio for example) and send that for the later registration process

There's a primitive instruction here - https://github.com/Yu-AnChen/palom#for-ome-tiff-files. Happy to discuss more. On the side, I also found wsireg (https://github.com/NHPatterson/wsireg) that is basically using very similar approach for aligning WSIs, I don't have experience with it but it might be helpful too.

alcrevenna commented 2 years ago

Dear @Yu-AnChen , thanks for the quick reply. I know about the link for open slide, I am in contact with Nico Curti, who is working on making vsi accessible via open slide. Indeed, hope full support will come soon.

I'll check WSireg and your instructions on PALOM. Somehow I got the impression that PALOM only worked on brightfield data, good to know it will work on fluorescence.

Yu-AnChen commented 2 years ago

Great! Hope either openslide or AICSimageio will have better support for vsi files.

And yes, palom works on fluorescent images too. Although there might be issues due to different imaging situations, e.g. imaging a whole slide vs an ROI. Happy to troubleshoot if things go wrong for you :)